P4P AEB issues

[Draken413o]

Hi all!

This is my first post to the forum. I've recently got the phantom 4 pro and have little problems but the main issue I'm facing is that AEB does not work well for me at all!
My problem is as follows, I set up the shot making my image have a balanced histogram with most of the details in the mid tones in the manual camera mode. (it also happens in aperture priority mode aswell.)

If I shoot a 3 AEB photo sequence, I get exposures that are 0, -1 and -0.3. This is pretty much useless to me in post processing!

If I were to shoot a 5 AEB photo sequence the issue gets more laughable, I would get an exposure sequence of 0, -2, -1, -2.3, -0.6

Even this is not consistent as I've noticed the exposure value change is not always the same from different flights

I'm not sure if I did anything wrong at all but how I'd normally set up for the shot would be in raw format, 3:2 crop, d-cinelike setting.

Also, I noticed when I view the raw images in lightroom CC 2015.8, I get the same settings for all the different exposures. This is quite strange as I expected it to have a variation in the shutter speed to correspond to the changes in exposure.

I would like to get my exposure sequence to be -4, -2, 0, +2, +4 if possible. Is there a setting to do this or is this a bug that other users are facing aswell?

I've uploaded some screenshots of my aeb sequence too!

I hope to hear from anyone else and thanks for reading!

 

[Meta4]

If I shoot a 3 AEB photo sequence, I get exposures that are 0, -1 and -0.3.

Are you sure about those numbers? I can't see the exposure details but it's usual for DJI to shoot 0, -0.7, +0.7.
Shooting 5AEB you should be getting -1.4, -0.7, 0, +0.7, +1.4

I would like to get my exposure sequence to be -4, -2, 0, +2, +4 if possible. Is there a setting to do this or is this a bug that other users are facing as well?

DJI have only ever allowed 0.7 stop increments but Litchi allows you to set a 1 stop increment.
So when Litchi comes out with a P4 pro compatible version, try that.

 

[Draken413o]

Are you sure about those numbers? I can't see the exposure details but it's usual for DJI to shoot 0, -0.7, +0.7.
Shooting 5AEB you should be getting -1.4, -0.7, 0, +0.7, +1.4


DJI have only ever allowed 0.7 stop increments but Litchi allows you to set a 1 stop increment.
So when Litchi comes out with a P4 pro compatible version, try that.

Hi Thanks for replying!

If assuming the earliest shot from the sequence is my 0 ev, the next shots are just getting darker for me. I'm not sure why.

I got the ev values from changing the exposure in lightroom of my first shot (once again assuming the first is the 0 ev image) to match the other images in the sequence to work out how much ev difference the aeb is making.

But at the end, I've noticed I do not have a consistent EV step size from various flights so working around this reliably is a little difficult.

 

[Meta4]

If assuming the earliest shot from the sequence is my 0 ev, the next shots are just getting darker for me. I'm not sure why.

Here's a bracket of 5 shot on a P3 and you can easily see the differences in the shutter speed.

My P4 pro seems to work exactly the same way.

 

[Draken413o]

Here's a bracket of 5 shot on a P3 and you can easily see the differences in the shutter speed.

My P4 pro seems to work exactly the same way.

So your P4P has a change in shutter speed in your AEB sequences? Mine doesn't. I tried it with RAW + Jpg and both file formats do not show differences in shutter speeds in the sequences.

 

[bkushner]

Are you sure about those numbers? I can't see the exposure details but it's usual for DJI to shoot 0, -0.7, +0.7.
Shooting 5AEB you should be getting -1.4, -0.7, 0, +0.7, +1.4


DJI have only ever allowed 0.7 stop increments but Litchi allows you to set a 1 stop increment.
So when Litchi comes out with a P4 pro compatible version, try that.

Yes before Litchi was working with P4P I was using the DJI app and noticed that .07 thingy..I would have liked at least -1 thankfully Litchi is working again (beta)

 

[Zzz]

So your P4P has a change in shutter speed in your AEB sequences? Mine doesn't. I tried it with RAW + Jpg and both file formats do not show differences in shutter speeds in the sequences.

My AEB metadata does not show any change in the camera settings either (ISO, aperture and shutter all supposedly remain the same throughout all 5 captures), just tells my by what fraction of a stop the exposure has been adjusted... so how do they achieve the said adjustment??

Was just about to create a thread about this myself when I stumbled upon yours...

 

[Meta4]

So your P4P has a change in shutter speed in your AEB sequences? Mine doesn't. I tried it with RAW + Jpg and both file formats do not show differences in shutter speeds in the sequences.

That was the older P3 Advanced, not the P4pro but it works the same except the P3 has no aperture to adjust.
Whether your camera adjusts the shutter speed or aperture will depend on the settings you have.
If you have the camera set to aperture priority, the aperture will be locked and the shutter speed will change ... and vice versa.

 

[Meta4]

My AEB meta data does not show any change in the camera settings either (ISO, aperture and shutter all supposedly remain remain the same throughout all 5 captures), just tells my by what fraction of a stop it's been adjusted... so how do they achieve the said adjustment??

Something had to change and it will be in the exif data ... look again.
If you have five photos with the same exif data, you were shooting burst mode rather than AEB.

 

[Zzz]

Something had to change and it will be in the exif data ... look again.
If you have five photos with the same exif data, you were shooting burst mode rather than AEB.

Nah, it's def AEB 5 mode, the exposure changes but the EXIF does not reflect how. Pretty sure the OP reported the same thing?

Here's a screenshot... the DNGs carry the same EXIF data as jpegs.

I was shooting full manual.

 

[Zzz]

And I think it's odd how the exposure adjustment sequence is not gradual but rather like this: 0, -0.7, +0.7, -1.3, +1.3

 

[Meta4]

If I shoot a 3 AEB photo sequence, I get exposures that are 0, -1 and -0.3. This is pretty much useless to me in post processing!
If I were to shoot a 5 AEB photo sequence the issue gets more laughable, I would get an exposure sequence of 0, -2, -1, -2.3, -0.6

I'd be extremely surprised if you are really getting those numbers.

Nah, it's def AEB 5 mode, the exposure changes but the EXIF does not reflect how. Pretty sure the OP reported the same thing?

Something's not right with the way you are viewing the exif data.
Here's a bracket of three I shot the other day on a P4 pro with exif data.
I was on aperture priority so the camera has changed the shutter speed.

 

[Zzz]

Not sure if you can see this cos embedding seems to downsize and bucher wider images...

 

[Zzz]

I've got some more info. Tested it tonight when it was dark and got some interesting results. Full manual including ISO 100, and while EXIF data doesn't show any differences, there is clear difference in the amount of noise between the pictures. ALL but the initial one set by me show noticeable amounts of noise, whether + or - EV, the initial one however is nearly noise free. Does that mean it adjusts exposure using ISO, but for some reason doesn't record that change in the EXIF? And does that mean it actually drops the ISO below 100 given that my initial set exposure at EV 0 uses ISO 100?

Thanks!

 

[Draken413o]

I've got some more info. Tested it tonight when it was dark and got some interesting results. Full manual including ISO 100, and while EXIF data doesn't show any differences, there is clear difference in the amount of noise between the pictures. ALL but the initial one set by me show noticeable amounts of noise, whether + or - EV, the initial one however is nearly noise free. Does that mean it adjusts exposure using ISO, but for some reason doesn't record that change in the EXIF? And does that mean it actually drops the ISO below 100 given that my initial set exposure at EV 0 uses ISO 100?

Thanks!

Yes I've noticed that too.. Also, I tried something interesting today aswell. If you shoot your aeb with the settings exposed to the right (google ETTR) and change the bring back the exposure in lightroom or whatever raw processor you are using, the result of the pulled back image is significantly better.

So.. the take away is for stills, if you want better image quality, shoot a little more over exposed than usual without clipping highlights and bring it back in post to have less noise in you shadows. I've used this technique for my canon dslr's and it seems to translate to this dji camera sensor aswell.

I've included a comparison zoomed at 1:1 crop. The image on the left was my 0 ev shot and the image on the right was from the aeb sequence that was second darkest.

I applied the same amount of noise reduction and sharpening to both. (sharpening 80, Noise reduction 20). Additionally to get a similar amount of noise between the ettr image, I would need to apply 60 noise reduction to the image on the right.

 

[Draken413o]

I just noticed something weird when looking at the properties. I've re arranged the images in terms of brightness and places the corresponding properties above the image. Apparently, my 0 step Exposure bias is the brightest image from my AEB 5 shot sequence.

By using my 0 exposure biased image and reducing the exposure to match the other images I got:

-0.7 exposure for the "+1.3 step" image
-1.0 exposure for the "+0.7 step" image
-2.1 exposure for the "-0.7 step" image
-2.43 exposure for the "-1.3 step" image

My AEB sequence is not working as it should be.

 

[phneeley]

This is pretty disappointing. While I can't comment on the metadata of my raw files (now deleted), I will say that my last batch of AEB photos were considerably less exposed than I was expecting. I ended up only using 1 of the sets because the rest were so bad.


Sent from my iPhone using PhantomPilots

 

[SmurfN]

so whats the conclusion? AEB doesn't really work (i.e provides noisy images) or just provides too small of a EV step with no adjustability to be worth the trouble?

 

[Max Tyler]

It's as if everybody is ignoring Meta4's post and going down a maddening worm hole.

 

[Rom3oDelta7]

First, I am not seeing any issues wrt the actual bracketed exposures. I did a bunch of bracketed exposures (5 shots) a few nights ago and the exposures all appear to be consistent with their EV adjustments. I just did a quick test, using:

1. shutter priority, Auto ISO
2. shutter priority, fixed ISO (set high enough that there was headroom for an f/stop adjustment)
3. aperture priority, fixed ISO (EV 0 is 1/100 shutter speed, so room to adjust down here also)
4. auto exposure, auto ISO

In all 4 sets of exposures, they are consistent with the EXIF file's indicated exposure bias: 0, -0.7, +0.7, -1.3, +1.3. This is apparent visually as well as observing the frequency shifts in the histogram in LR 2015.8. So it appears to be bracketing correctly but not documenting what it adjusted in the EXIF data. To adjust the exposure, one of the parameters of ISO, aperture, or shutter speed must be altered (the exposure triangle), so something is definitely amiss in the EXIF data (whether viewed in Win10 or LR's metadata window). I checked some online EXIF viewers and they were consistent with this, so I could not reproduce @Meta4's results.

Another thread is talking about how the timestamp on files can be wrong depending on the size of the SD card; not sure if that has any bearing here, but this was with a 64GB card.